RAG Engineer — Cohere India

About this RAG Engineer role at Cohere

Cohere, a leader in generative AI and LLM technology, is building its India operations and seeking RAG Engineers in Bengaluru. This is a mid-level position for experienced AI engineers who want to work on retrieval-augmented generation systems at scale—a core competency in modern LLM applications. Based in Bengaluru with hybrid flexibility, this role puts you at the center of AI infrastructure innovation.

What you'll do

• Design and optimize retrieval-augmented generation (RAG) pipelines that combine LLMs with external knowledge sources for improved accuracy and relevance.
• Work with vector databases like Weaviate, Qdrant, or Pinecone to manage embeddings and implement efficient semantic search at scale.
• Develop and fine-tune embedding models that power retrieval quality, ensuring relevance and performance in production environments.
• Build LLM application architectures that integrate retrieval systems, prompting strategies, and response generation into cohesive production systems.
• Deploy and monitor RAG systems in production, managing performance, latency, and cost optimization across real-world workloads.
• Collaborate with cross-functional teams to implement information retrieval best practices and improve model performance iteratively.

Who should apply

You need 3+ years of hands-on experience building retrieval-augmented systems or similar AI infrastructure. This is a mid-level role—not entry-level—so you should have shipped production AI systems before. Deep expertise in embeddings and vector databases is essential; familiarity with at least one of Weaviate, Qdrant, or Pinecone is expected. You understand LLM application architecture and deployment challenges. If you're a senior engineer with this background, you're absolutely welcome to apply; Cohere values strong technical depth.

Salary & offer in context

Cohere offers ₹40–80 LPA (₹3,33,333–₹6,66,667 per month), positioning this role in the upper-mid band for AI engineer salaries in India. For a RAG Engineer in Bengaluru with 3+ years experience at a private MNC, this reflects market rates for specialized AI infrastructure roles—particularly at a company betting heavily on LLM capabilities.

Path to apply

This is a direct online application through Cohere's careers portal—no exam, no gatekeeping process. You apply directly to Cohere and move through their standard technical interview process. The application deadline is 13 June 2026, so if this role interests you, apply within the next two weeks. Cohere typically moves quickly on hiring for India operations.